A Temporal Fusion Approach for Video Classification with Convolutional and LSTM Neural Networks Applied to Violence Detection

In modern smart cities, there is a quest for the highest level of integration and automation service. In the surveillance sector, one of the main challenges is to automate the analysis of videos in real-time to identify critical situations. This paper presents intelligent models based on Convolutio...

Full description

Bibliographic Details
Main Authors:	Jean Phelipe de Oliveira Lima, Carlos Maurício Seródio Figueiredo
Format:	Article
Language:	English
Published:	Asociación Española para la Inteligencia Artificial 2021-02-01
Series:	Inteligencia Artificial
Subjects:	Applications of AI Deep Learning Intelligent Video Processing Violence Detection
Online Access:	https://journal.iberamia.org/index.php/intartif/article/view/573

id	doaj-2cd6e88732b945c7ac7e2c1f86597b26
record_format	Article
spelling	doaj-2cd6e88732b945c7ac7e2c1f86597b262021-03-07T01:11:13ZengAsociación Española para la Inteligencia ArtificialInteligencia Artificial1137-36011988-30642021-02-01246710.4114/intartif.vol24iss67pp40-50A Temporal Fusion Approach for Video Classification with Convolutional and LSTM Neural Networks Applied to Violence DetectionJean Phelipe de Oliveira Lima 0Carlos Maurício Seródio Figueiredo1Universidade do Estado do Amazonas, BrazilUniversidade do Estado do Amazonas, Brazil In modern smart cities, there is a quest for the highest level of integration and automation service. In the surveillance sector, one of the main challenges is to automate the analysis of videos in real-time to identify critical situations. This paper presents intelligent models based on Convolutional Neural Networks (in which the MobileNet, InceptionV3 and VGG16 networks had used), LSTM networks and feedforward networks for the task of classifying videos under the classes "Violence" and "Non-Violence", using for this the RLVS database. Different data representations held used according to the Temporal Fusion techniques. The best outcome achieved was Accuracy and F1-Score of 0.91, a higher result compared to those found in similar researches for works conducted on the same database. https://journal.iberamia.org/index.php/intartif/article/view/573Applications of AIDeep LearningIntelligent Video ProcessingViolence Detection
collection	DOAJ
language	English
format	Article
sources	DOAJ
author	Jean Phelipe de Oliveira Lima Carlos Maurício Seródio Figueiredo
spellingShingle	Jean Phelipe de Oliveira Lima Carlos Maurício Seródio Figueiredo A Temporal Fusion Approach for Video Classification with Convolutional and LSTM Neural Networks Applied to Violence Detection Inteligencia Artificial Applications of AI Deep Learning Intelligent Video Processing Violence Detection
author_facet	Jean Phelipe de Oliveira Lima Carlos Maurício Seródio Figueiredo
author_sort	Jean Phelipe de Oliveira Lima
title	A Temporal Fusion Approach for Video Classification with Convolutional and LSTM Neural Networks Applied to Violence Detection
title_short	A Temporal Fusion Approach for Video Classification with Convolutional and LSTM Neural Networks Applied to Violence Detection
title_full	A Temporal Fusion Approach for Video Classification with Convolutional and LSTM Neural Networks Applied to Violence Detection
title_fullStr	A Temporal Fusion Approach for Video Classification with Convolutional and LSTM Neural Networks Applied to Violence Detection
title_full_unstemmed	A Temporal Fusion Approach for Video Classification with Convolutional and LSTM Neural Networks Applied to Violence Detection
title_sort	temporal fusion approach for video classification with convolutional and lstm neural networks applied to violence detection
publisher	Asociación Española para la Inteligencia Artificial
series	Inteligencia Artificial
issn	1137-3601 1988-3064
publishDate	2021-02-01
description	In modern smart cities, there is a quest for the highest level of integration and automation service. In the surveillance sector, one of the main challenges is to automate the analysis of videos in real-time to identify critical situations. This paper presents intelligent models based on Convolutional Neural Networks (in which the MobileNet, InceptionV3 and VGG16 networks had used), LSTM networks and feedforward networks for the task of classifying videos under the classes "Violence" and "Non-Violence", using for this the RLVS database. Different data representations held used according to the Temporal Fusion techniques. The best outcome achieved was Accuracy and F1-Score of 0.91, a higher result compared to those found in similar researches for works conducted on the same database.
topic	Applications of AI Deep Learning Intelligent Video Processing Violence Detection
url	https://journal.iberamia.org/index.php/intartif/article/view/573
work_keys_str_mv	AT jeanphelipedeoliveiralima atemporalfusionapproachforvideoclassificationwithconvolutionalandlstmneuralnetworksappliedtoviolencedetection AT carlosmauricioserodiofigueiredo atemporalfusionapproachforvideoclassificationwithconvolutionalandlstmneuralnetworksappliedtoviolencedetection AT jeanphelipedeoliveiralima temporalfusionapproachforvideoclassificationwithconvolutionalandlstmneuralnetworksappliedtoviolencedetection AT carlosmauricioserodiofigueiredo temporalfusionapproachforvideoclassificationwithconvolutionalandlstmneuralnetworksappliedtoviolencedetection
_version_	1724229488627154944

A Temporal Fusion Approach for Video Classification with Convolutional and LSTM Neural Networks Applied to Violence Detection

Similar Items